1 Marginal Satisfaction of Recreational Hunters Red Deer Harvests Geoffrey N. Kerr Lincoln University Contributed paper prepared for presentation at the 61st AARES Annual Conference, Brisbane Conference and Entertainment Centre, Brisbane, 8-10 February 2017. Copyright 2017 by Geoffrey Kerr. All rights reserved. Readers may make verbatim copies of this document for non-commercial purposes by any means, provided that this copyright notice appears on all such copies.
Marginal satisfaction of recreational hunters red deer harvests Geoff Kerr Australian Agricultural and Resource Economics Society annual conference, Brisbane 7-10 February 2017
Game Animal Council Act 2013 Deer were formerly pests Open access resource Competition to harvest No incentives to invest in quantity or quality Low deer numbers Poor trophy prospects Low value of recreational hunting Deer are now game that can be managed as a hunting resource What are the costs and benefits of management? How can management add value?
Managing for Satisfaction Would a bag limit increase aggregate hunter satisfaction? What makes a hunt satisfying? Seeing game, killing game, social, scenery, etc. What role do motivations play? Is motivation important per se? Research question Is more better? Gossen s Law implies equal marginal utilities for efficiency, does it apply? Does marginal utility diminish significantly with harvest? Photo: Geoff Kerr
Seeing & killing game: prior studies, different contexts 1980s Wildlife Society Bulletin Decker, D. et al. (1980). Further insights into the multiple-satisfactions approach for hunter management. Wildlife Society Bulletin 8: 323-331. McCullough, D.R. & Carmen, W.J. (1982). Management goals for deer hunter satisfaction. Wildlife Society Bulletin 10: 49-52. Rollins, R. & Romano, L. (1989). Hunter satisfaction with the selective harvest system for moose management in Ontario. Wildlife Society Bulletin 17: 470-475. Hammitt, W.E., et al. (1990). Determinants of multiple satisfaction for deer hunting. Wildlife Society Bulletin 18: 331-337. 2000s Human Dimensions of Wildlife Gigliotti, L. (2000). A classification scheme to better understand satisfaction of Black Hills deer hunters: The role of harvest success. Human Dimensions of Wildlife 5: 32-51. Heberlein, T.A. & Kuentzel, W.F. (2002). Too many hunters or not enough deer? Human and biological determinants of hunter satisfaction and quality. Human Dimensions of Wildlife 7: 229-250. Frey, S.N., et al. (2003). Factors influencing pheasant hunter harvest and satisfaction. Human Dimensions of Wildlife 8: 277-286. Fulton, D.C. & Manfredo, M.J. (2004). A panel design to assess the effects of regulatory induced reductions in opportunity on deer hunters satisfaction. Human Dimensions of Wildlife 9: 35-55.
Data Tracked ~900 big game hunters for a year (June 2011/June 2012) Chose one hunt at random each month Measured specific hunt motivations, sightings, kills Measured satisfaction on a single item scale Rate this hunt: (1) very unsatisfied to (5) very satisfied A measure of overall satisfaction Already had information on demographics, motivations for hunting in general, relative importance of aspects of hunting, and hunting avidity Data on 2,917 red deer hunts by 698 different hunters
Data: 698 hunters Variable N Mean SD Median Age 697 39.74 13.05 40 Years of big game hunting experience 696 22.04 14.36 21 Days spent big game hunting per year 697 32.66 29.63 25 Big game hunts per year 696 16.97 21.53 12 Red deer killed per year 531 3.09 5.13 2 Male 698 97.9% Maori 698 8.3% North Island resident 698 50.1% NZ Deerstalkers Association member 698 35.0% Primary motivation to hunt: Enjoy outdoors 698 50.0% Primary motivation to hunt: Meat 698 19.1% Primary motivation to hunt: See wild animals 698 7.2% Primary motivation to hunt: Excitement 698 6.6% Primary motivation to hunt: Get away from civilisation 698 6.2% Primary motivation to hunt: Trophy 698 5.6%
Data: 2,917 hunts Variable N Mean SD Median One way travel distance (km) 2910 136.95 184.01 80 One way travel time (hours) 2910 3.23 9.27 1.5 Cost of travel (NZ$) 2912 118.87 238.35 50 Days hunted 2909 2.16 2.00 1 Number of hunters in the party 2912 2.07 1.17 2 Number of red deer the individual killed 2763 0.44 0.88 0 Primary motivation for this hunt: Enjoy outdoors 2917 33.5% Primary motivation for this hunt: Meat 2917 29.4% Primary motivation for this hunt: Trophy 2917 10.9% Saw red deer 2917 64.0% Didn t kill a red deer 2763 68.2% Killed 1 red deer 2763 23.7% Killed 2 red deer 2763 6.0% Killed 3 or more red deer 2763 2.1% Didn t kill a red deer, but another party member did 2756 10.0%
Killing deer affects satisfaction 2 = 287.7, P <<.0001 70% Satisfaction by red deer killed Very satisfied hunters are more likely to have killed a deer Not much difference if more than 1 kill 60% 50% 40% 30% 20% 10% 0% No deer killed 1 deer killed At least 2 deer killed Not satisfied Satisfied Very satisfied Photo: Geoff Kerr
Data analysis Latent class ordered logit model Preference heterogeneity is endemic Latent class better than RPL Accounts for correlated responses in panel data and for multivariate correlations Similar to Frey, S.N., et al. (2003). Factors influencing pheasant hunter harvest and satisfaction. Human Dimensions of Wildlife 8: 277-286. Photo: Jamie Carle
Latent class ordered logit model Parameter Class 1 Class 2 Class 3 Class 4 Constant 1.49 *** 2.05 *** -0.31-1.38 *** Category threshold 1.62 *** 3.38 *** 4.57 *** 1.66 *** Saw a red deer 1.33 *** 0.22 1.43 *** 1.01 *** Killed one red deer 0.71 0.91 ** 4.59 *** 0.77 ** Killed two or more red deer 0.29 1.67 *** -1.09-0.11 Individual did not kill red deer, but party did 0.64 0.07 5.71 *** 0.62 * Meat hunt -1.18 *** -0.85 ** -1.37 ** -1.07 ** Interaction: Meat hunt x killed a red deer 1.52 *** -1.16 ** 0.38 1.97 *** NZ Deerstalkers Association member 0.25 0.21 0.09 0.87 *** Class probability 0.31 *** 0.23 *** 0.19 *** 0.26 *** N hunts 2756 N hunters 698 * <.10 LL (restricted) -2984.141 ** <.05 LL (model) -2527.055 *** <.01 BIC/N 1.946 abic/n 1.901 Adjusted Rho 2 0.140
Probability Probability Marginal effects, first deer killed 1.000 1.000 0.800 0.800 0.600 0.600 0.400 0.200 0.000-0.200 0.400 0.200 0.000-0.200-0.400-0.400-0.600-0.600-0.800-0.800 P0 P1 P2 P0 P1 P2 P0 P1 P2 P0 P1 P2-1.000 P0 P1 P2 P0 P1 P2 P0 P1 P2 P0 P1 P2 Class 1 Class 2 Class 3 Class 4 Class 1 Class 2 Class 3 Class 4 Non-meat hunt Meat hunt Non-meat hunt: Significant positive effects for Classes 2, 3 and 4 Meat hunt: Significant positive effects for Classes 1, 3 and 4 P0 Probability not satisfied P1 Probability satisfied P2 Probability very satisfied
Are two kills better than one? Photo: Geoff Kerr Photo: Geoff Kerr
Probability Probability Marginal effects, second deer killed 0.600 0.400 0.200 1.000 0.800 0.600 0.400 0.000 0.200 0.000-0.200-0.200-0.400-0.400-0.600-0.600 P0 P1 P2 P0 P1 P2 P0 P1 P2 P0 P1 P2-0.800 P0 P1 P2 P0 P1 P2 P0 P1 P2 P0 P1 P2 Class 1 Class 2 Class 3 Class 4 Class 1 Class 2 Class 3 Class 4 Non-meat hunt Meat hunt Non-meat hunt: Effects not significant Meat hunt: Significant positive effects for Class 2 (positive) and Class 4(negative)
Why a second kill might be sub-optimal for a meat hunter! Photo: Hendrik Venter
Nonmeat hunt Marginal utility 10,000 Monte Carlo replications Class 1 Class 2 Class 3 Class 4 Nonmeat Non- Non- Meat Meat meat Meat meat hunt hunt hunt hunt hunt hunt Meat hunt MU1 0.71 2.24 *** 0.91 ** -0.26 4.59 *** 4.97 *** 0.77 ** 2.73 *** MU2 0.29 0.29 1.67 *** 1.67 *** -1.09-1.09-0.11-0.11 MU1-MU2 0.42 1.95 * -0.77-1.94 * 5.68 *** 6.06 ** 0.87 * 2.84 *** Diminishing MU? X MU1 Marginal utility from first kill MU2 Marginal utility from second kill * = 0.10; ** = 0.05; *** = 0.01
Class 2 hunters are little different to others Mean (SEM) Class 1 Class 2 Class 3 Class 4 F Sig Annual game hunts 18.43 16.96 17.18 15.13 0.748.524 Annual days game hunting 34.70 31.89 32.85 30.92 0.583.626 NZDA member 0.35 0.36 0.30 0.37 0.514.673 Experience (years) 23.19 22.42 18.98 22.42 2.320.074 Age (years) 40.92 3 40.38 3 36.53 1,2 39.80 3.116.026 Importance of killing game 1.78 1.96 1.94 1.83 2.437.064 Importance of trophy 1.67 1.80 1.73 1.81 1.267.285 Importance of harvesting meat 2.53 2.47 2.53 2.51 0.233.873 Main reason to hunt is meat 0.20 0.19 0.17 0.19 0.145.933 Main reason to hunt is trophy 0.05 0.08 0.03 0.07 1.257.288 Annual red deer harvest 2.83 3.29 3.28 3.06 0.213.888 Killed one deer this hunt 0.26 3 0.24 0.19 1 0.24 2.912.033 Killed 2 or more deer this hunt 0.09 0.08 0.07 0.08 0.481.695 This hunt was a meat hunt 0.32 0.29 0.27 0.29 1.253.289 This hunt was a trophy hunt 0.09 2 0.14 1,4 0.11 0.09 2 4.505.004 Numbered superscripts identify group mean differences using Tukey HSD test at P.05
Allocation evaluation Comparison of coefficients across classes is not valid Cardinality of satisfaction can not be invoked Problem for optimisation across classes Assume: Deer kills are reallocated within each category of hunter (Class x Motive) i.e. the 298 deer killed in 624 non-meat hunts by class 1 hunters are reallocated across those 624 hunts This is a strong assumption, but it permits evaluation of a within group one deer bag limit Assuming reallocation within classes (ignore motive) would permit within class optimisation
Class 1 Non-meat hunts N = 624 One deer bag limit Hunts on which no deer are killed Hunts on which 1 deer is killed Hunts on which 2 or more deer are killed Total Utility per hunt 0 0.712 1.000 Observed hunts 430 143 51 298 Deer Utility observed 0 101.82 51.04 152.86 Hunts under one deer bag limit 326 298 0 298 Deer Utility bag limit 0 212.19 0 212.19 Utility 0 110.37-51.04 59.32 Z( Utility = 0) 0.942 p(z) 0.346 All utility estimates are relative, U(zero kills) = 0 by assumption. Z from 10,000 Monte Carlo simulations
What about other classes? 10,000 Monte Carlo utility difference simulations Reported statistics are (1) sign of utility change and (2) p(z) mean utility change = zero Class 1 Class 2 Class 3 Class 4 Non-meat hunt + p=.346 - p=.780 + p=.008 + p=.051 Meat hunt + p=.047 - p=.077 + p=.010 + p=.000
Management Implications Multiple kills are not a big issue: Only 8% of hunts have multiple kills, 24% kill one deer, 68% don t kill a deer But, in most cases the first kill is very valuable Few would be harmed by a one deer bag limit (Maybe Class 2 meat hunters) But Class 3 and 4 hunters would definitely benefit Meat hunts are different for everybody Meat hunters need to kill to be as satisfied as non-meat hunters without a kill
Management Implications Efficient to reallocate second kills from Class 4 to Class 2 MS2 Class 4 < 0, MS2 Class 2 = >0 But why do Class 4 hunters kill the second deer? Are hunters ex ante and ex post perspectives different? More research needed Efficient allocation requires differentiation of class membership and motivations This is not possible by observing external characteristics
Conclusions A simple overall satisfaction question enabled identification of relative values of attributes of the hunting experience Diminishing marginal utility (Gossen s Law) may not apply to all hunters Hunter heterogeneity implies the need to manage some hunters differently Since class membership is not observable, it is not clear how that could be done Overall, a one deer bag limit is likely to have efficiency benefits
Photo: Graham Nugent geoffrey.kerr@lincoln.ac.nz